The broad, long-term objective of the proposed research Is to provide Insight Into how the BER and RAD51 variants In the normal population and In tumors contribute to altered DNA repair capacity, Altered DNA repair capacity can drive carcinogenesis and Is Important for designing personalized cancer therapy. The specific aims are: (1) to test the hypothesis that variants of NTHI, NEIL1, 2, 3, and RAD51 that are found in the normal population and In tumors are linked to cancer; (2) to test the hypothesis that DNA glycosylase and RAD51 variants arising in the normal population and In tumors lead to genomic instability; and (3) to test the hypothesis that DNA repair variants Influence cellular responses to DNA damaging agents. To accomplish these alms we will express the variants in tissue culture cells and determine if they Induce cellular transformation. We will also Initiate studies to Identify additional tumor-associated variants within our collection of DNA samples Isolated from colon and breast carcinomas. We will measure the in vivo mutation frequency and generate mutation spectra to determine the types of mutations that result from cells expressing variant DNA glycosylases and Rad51. In cells expressing these variants we will also characterize responses to ionizing radiation, cisplatin, and other appropriate agents to determine if they Influence DNA repair. Information gleaned from Project 1 will guide additional biochemical and structural studies on the variants In Projects 2 and 3 as well as their Interactions with nucleosomes In Project 4. Core A will provide the bioinformatics and statistical Infrastructure for Project 1 and Core B will provide protein constructs and enzymes.